Models and providers
N4XION AI gives you one assistant called OPAW v1. You pick a mode and send a request. OPAW handles the rest. You see the mode and whether it is available, not the raw provider that runs behind it.
OPAW only uses free endpoints. FREE_ONLY is on, so paid endpoints are never used. When you leave the mode on Auto, OPAW picks one suitable free provider for your request and uses a fallback if the first one does not respond. It never calls every provider for a single request.
Providers OPAW can route across
When an administrator has enabled a provider and verified a free model, OPAW can route your request across these free providers.
- Google Gemini
- Groq
- Mistral
- OpenRouter, free models only
- NVIDIA
- Cerebras
- SambaNova
- GitHub Models
- Cloudflare Workers AI
Web mode search uses Tavily. OPAW searches with Tavily first, then a model writes the answer.
Availability can change
Which providers are on depends on what an administrator has enabled and verified. A free provider can hit its own limit, so OPAW may switch to another provider, and a limit can pause a mode for a while. Because of this, the exact model that answers your request can change from one request to the next.
Image, Video, and Voice are currently unavailable. They show an unavailable message until a configured free endpoint can support them.
Questions
If a mode is not working the way you expect, email hello@n4xion.net. You can also read the Privacy Policy to see how your requests are handled.